Poly( N , N -dimethylacrylamide)-Coated Maghemite Nanoparticles for Stem Cell Labeling

Subscriber access provided by CZECH ACADEMY OF SCIENCES

Bioconjugate Chemistry is published by the American Chemical Society. 1155Sixteenth Street N.W., Washington, DC 20036

Article

Poly(N,N-dimethylacrylamide)-CoatedMaghemite Nanoparticles for Stem Cell Labeling

Michal Babic#, Daniel Hora#k, Pavla Jendelova#, Kater#ina Glogarova#, Vi#t Herynek, MiroslavaTrchova#, Katari#na Likavc#anova#, Petr Lesny#, Emil Pollert, Milan Ha#jek, and Eva Sykova#

Bioconjugate Chem., 2009, 20 (2), 283-294• DOI: 10.1021/bc800373x • Publication Date (Web): 14 January 2009

Downloaded from http://pubs.acs.org on February 18, 2009

More About This Article

Additional resources and features associated with this article are available within the HTML version:

• Supporting Information• Access to high resolution figures• Links to articles and content related to this article• Copyright permission to reproduce figures and/or text from this article

http://pubs.acs.org/doi/full/10.1021/bc800373x

Poly(N,N-dimethylacrylamide)-Coated Maghemite Nanoparticles for StemCell Labeling

Michal Babic,†,‡ Daniel Horak,*,†,‡ Pavla Jendelova,‡,§ Katerina Glogarova,§ Vıt Herynek,‡,| Miroslava Trchova,†

Katarına Likavcanova,§ Petr Lesny,‡,§ Emil Pollert,⊥ Milan Hajek,‡,| and Eva Sykova‡,§

Institute of Macromolecular Chemistry, Academy of Sciences of the Czech Republic, Heyrovsky Sq. 2, 162 06 Prague 6,Czech Republic, Center for Cell Therapy and Tissue Repair, Charles University, V Uvalu 84, 150 06 Prague 5, Czech Republic,Institute of Experimental Medicine, Academy of Sciences of the Czech Republic, Vıdenska 1083, 142 20 Prague 4, CzechRepublic, Institute for Clinical and Experimental Medicine, Vıdenska 1958/9, 140 21 Prague 4, Czech Republic, and Institute ofPhysics, Academy of Sciences of the Czech Republic, Cukrovarnicka 10, 162 53 Prague 6, Czech Republic. ReceivedSeptember 3, 2008; Revised Manuscript Received November 13, 2008

Maghemite (γ-Fe2O3) nanoparticles were obtained by the coprecipitation of Fe(II) and Fe (III) salts with ammoniumhydroxide followed by oxidation with sodium hypochlorite. Solution radical polymerization of N,N-dimethyl-acrylamide (DMAAm) in the presence of maghemite nanoparticles yielded poly(N,N-dimethylacrylamide)(PDMAAm)-coated maghemite nanoparticles. The presence of PDMAAm on the maghemite particle surface wasconfirmed by elemental analysis and ATR FTIR spectroscopy. Other methods of nanoparticle characterizationinvolved scanning and transmission electron microscopy, atomic adsorption spectroscopy (AAS), and dynamiclight scattering (DLS). The conversion of DMAAm during polymerization and the molecular weight of PDMAAmbound to maghemite were determined by using gas and size-exclusion chromatography, respectively. The effectof ionic 4,4′-azobis(4-cyanovaleric acid) (ACVA) initiator on nanoparticle morphology was elucidated. Thenanoparticles exhibited long-term colloidal stability in water or physiological buffer. Rat and human bone marrowmesenchymal stem cells (MSCs) were labeled with uncoated and PDMAAm-coated maghemite nanoparticlesand with Endorem as a control. Uptake of the nanoparticles was evaluated by Prussian Blue staining, transmissionelectron microscopy, T2-MR relaxometry, and iron content analysis. Significant differences in labeling efficiencywere found for human and rat cells. PDMAAm-modified nanoparticles demonstrated a higher efficiency ofintracellular uptake into human cells in comparison with that of dextran-modified (Endorem) and unmodifiednanoparticles. In gelatin, even a small number of labeled cells changed the contrast in MR images. PDMAAm-coated nanoparticles provided the highest T2 relaxivity of all the investigated particles. In vivo MR imaging ofPDMAAm-modified iron oxide-labeled rMSCs implanted in a rat brain confirmed their better resolution comparedwith Endorem-labeled cells.

INTRODUCTION

Recently, superparamagnetic iron oxide nanoparticles havefound widespread application in medicine, in particular ascontrast agents in magnetic resonance imaging (MRI) (1).Contrast agents in MRI improve the detection and characteriza-tion of lesions by changing the signal intensity of target tissueswhen compared with the surrounding tissue (2). Understandingthe interactions between the nanoparticles and cells is crucialfor improving their behavior in vivo and in vitro. The directuse of nanoparticles as in vivo MRI contrast agents results intheir biofouling in the blood plasma and the formation ofaggregates that are quickly sequestered by cells of the reticularendothelial system (RES) such as macrophages. Both smallparticle size and surface chemical structures are key parametersdetermining the blood half-life in the circulation, opsonization,biokinetics,andthebiodistributionofmagneticnanoparticles(3–5).

It is therefore essential to engineer the surface of the iron oxideto minimize biofouling and the aggregation of the particles underphysicochemical conditions (i.e., high salt and protein concen-trations) for long periods. Surface modifications of magneticiron oxide nanoparticles with biocompatible polymers arepotentially beneficial for preparing MR contrast agents for invivo applications. Such particles can circulate in the plasma forlong periods by escaping uptake by the RES. Iron oxidenanoparticles coated with dextran and its derivates, such asFeridex (Endorem) and Resovist, are commercially available.Iron oxide nanoparticles also show potential for in vitro celllabeling followed by in vivo MRI tracking, an approach that isbeing investigated for stem cell therapy (6, 7). Dextran-coatedagents are, however, insufficiently taken up by the cells duringcultivation; consequently, modifications of the iron oxide surfaceare attempted in order to increase the internalization of thenanoparticles. One promising approach toward increasing the lo-cal concentration of the nanoparticles in cells is to conjugatethem with targeting molecules that have a high affinity for tumorcells (8–10). Target-specific molecules, including monoclonalantibodies (11), proteins (12), peptides, and folic acid (10), havebeen investigated to increase the site-specific accumulation ofMR contrast agents in tumor cells. Several transfection agentshave been suggested to facilitate the effective magnetic labelingof cells by endocytosis into endosomes (13). These are mostlycationic, positively charged compounds (14). Superparamagnetic

* To whom correspondence should be addressed. E-mail: [email protected].

† Institute of Macromolecular Chemistry, Academy of Sciences ofthe Czech Republic.

‡ Charles University.§ Institute of Experimental Medicine, Academy of Sciences of the

Czech Republic.| Institute for Clinical and Experimental Medicine.⊥ Institute of Physics, Academy of Sciences of the Czech Republic.

Bioconjugate Chem. 2009, 20, 283–294 283

10.1021/bc800373x CCC: $40.75 2009 American Chemical SocietyPublished on Web 01/14/2009

iron oxide nanoparticles have thus been incorporated intohemoglobin within red cells, short human immunodeficiencyvirus-transactivator transcription (TAT) proteins (15), polyamines(polycationic poly(L-lysine) (PLL) (16)), magnetodendrimers(17), and antigen-specific internalizing monoclonal antibodies(18). The complexing of PLL with iron oxide occurs throughelectrostatic interactions of the negatively charged iron oxidenanoparticles with the polycation, allowing for the efficientincorporation of iron oxide nanoparticles into endosomes formagnetic cell labeling. A disadvantage of PLL-modified ironoxides is, however, that they aggregate in the presence of someelectrolytes.

In our previous reports, the efficiency of cellular uptake wasinvestigated following D-mannose and PLL modification of theiron oxide surface (19, 20). These systems were relatively stablein cell culture media; however, their further modification withantibodies, enzymes, or proteins, was difficult. In an attempt todesign a novel coating of iron oxide nanoparticles that wouldnot only ensure their very good colloidal stability but also enablethe subsequent covalent attachment of highly specific biomol-ecules (antibodies, enzymes, peptides, etc.), PDMAAm wasselected. Its advantage consists in the easy formation ofhomogeneous networks and copolymers with various functionalmonomers (21). Our preliminary biological experiments con-firmed not only the nontoxicity of the PDMAAm coating butalso its ability to enhance cellular uptake. In this study,PDMAAm coating of iron oxide particles was investigated inmore detail utilizing in vitro cell labeling followed by MR celltracking. As a model, rat and human mesenchymal stem cells(rMSCs and hMSCs) were labeled with PDMAAm-coatednanoparticles, and nanoparticle internalization into cells wasvisualized with light and TEM microscopy. In addition, theeffect of PDMAAm-coated iron oxide nanoparticles on cellviability and labeling efficiency was quantified. Moreover, MRrelaxometry of labeled cells was used to measure relaxationtimes, which are responsible for contrast enhancement in MRimages.

EXPERIMENTAL SECTION

Materials. N,N-Dimethylacrylamide (DMAAm; Aldrich,Milwaukee, USA) was distilled (56 °C/266 Pa). 4,4′-Azobis(4-cyanovaleric acid) (ACVA), FeCl2 ·4H2O, and FeCl3 ·6H2O werepurchased from Fluka (Buchs, Switzerland) and used withoutfurther purification. All other chemicals were supplied byAldrich and used as received. Endorem, a MRI contrast agentbased on superparamagnetic dextran-coated iron oxide nano-particles, was obtained from Guerbet (Roissy, France). UltrapureQ water ultrafiltered on a Milli-Q Gradient A10 system(Millipore, Molsheim, France) was used for preparation of thesolutions.

Preparation of Ferrofluid. Two procedures are generallyused for the preparation of iron oxide nanoparticles: an in situand a stepwise method. In this report, a two-step procedure wasused. Colloidal Fe(OH)3 was first precipitated from FeCl3 ·6H2Oadded to less than an equimolar amount of ammonia, followedby the addition of FeCl2 ·4H2O (molar ratio Fe(III)/Fe(II) ) 2).Mixing was simply achieved by rapid sonication. The mixturewas then poured into an excess of ammonia, and a magnetite(Fe3O4) coagulate was formed. The key step in the formationof a stable magnetite colloid after the synthesis consists in itscareful purification from all impurities by washing and magneticseparation in water, a process called peptization. Briefly, 12 mL0.2 M FeCl3 ·6H2O aqueous solution was mixed with 12 mL0.5 M NH4OH solution under sonication (Sonicator W-385; HeatSystems-Ultrasonics, Inc., Farmingdale, USA) for 2 min atlaboratory temperature to form the Fe(OH)3 colloid. Then, 6mL of 0.2 M FeCl2 ·4H2O aqueous solution was added under

sonication and the mixture poured into 36 mL of 0.5 M NH4OHaqueous solution. The resulting magnetite coagulate was left togrow for 15 min, magnetically separated, and repeatedly(7-10×) washed with Q-water to remove all impurities(including NH4Cl) remaining after the synthesis. The above-formed pure magnetite was sonicated for 3 min with 2 mL of0.1 M sodium citrate solution and oxidized at room temperaturewith 1.5 mL of 5 wt % sodium hypochlorite to maghemite. Theprecipitate was again repeatedly (7×) washed with Q-waterusing magnetic separation/decantation to obtain a stable colloid.

Coating by Solution Radical Polymerization ofDMAAm in the Presence of Maghemite Nanoparticles. In atypical experiment, 1 g of DMAAm and 0.01 g of ACVAinitiator were dissolved in 30 mL of aqueous colloid containing0.5 g of γ-Fe2O3, and the mixture was purged with nitrogen toremove oxygen for 10 min. PDMAAm was synthesized understirring (400 rpm) at 70 °C for 8 h. The resulting colloid waswashed 8 times with water by centrifugation/redispersion toremove excessive polymer, unreacted monomer, and initiator,and redispersed.

For comparison, iron oxide nanoparticles were also encap-sulated in neat PDMAAm added to the particles, followed by athorough purification by repetitive centrifugation/redispersionin water.

Characterization of the Particles. The particles wereobserved in a scanning electron microscope (SEM, JEOL JSM6400). A drop of particle dispersion was smeared betweenglasses for microscopy to obtain a thin layer, dried, and surface-coated with a 4-nm platinum layer using a SCD 050 sputteringdevice (BalTech, Lichtenstein). Transmission electron micros-copy (TEM) was performed on a JEOL JEM 200 CX todetermine the particle size of the iron oxides. The averagediameter and particle size distribution PDI (weight-number-average particle diameter ratio Dw/Dn) were obtained bystatistical analysis of at least 500 particles using the Atlasprogram (Tescan Brno, Czech Republic). The X-ray diffractionmeasurement was carried out with a Bruker D8 diffractometerusing Cu KR radiation and a Sol-X energy dispersive detector.The amount of iron in the coated nanoparticles was found byAAS (Perkin-Elmer 3110) of an extract of the sample obtainedwith dilute HCl (1:1) at 80 °C for 1 h. The amount of PDMAAmbound to the particles after polymerization was determined aftertheir lyophilization by elemental analysis (Perkin-Elmer 2400Series II CHNS/O Analyzer, Shelton, CT, USA).

The coating of PDMAAm on the surface of magneticnanoparticles was investigated using a Thermo Nicolet NEXUS870 FTIR Spectrometer (Madison, WI, USA) in a water-purgedenvironment with a DTGS detector in the wavenumber rangefrom 400 to 4000 cm-1. A Golden Gate Heated Diamond ATRTop-Plate (MKII Golden Gate single reflection ATR system)(Specac Ltd., Orprington, Great Britain) was used for measuringthe spectra of powdered samples by ATR spectroscopy. Typicalparameters used were 256 sample scans, 4 cm-1 resolution,Happ-Genzel apodization, and KBr beamsplitter.

The hydrodynamic diameter Dh and the zeta-potential of boththe unwashed colloid (0.2 mL of original colloid diluted to 2mL) and the colloid repeatedly (8×) washed with water throughultracentrifugation/redispersion were determined by dynamiclight scattering (DLS) with an Autosizer Lo-C (MalvernInstruments Ltd., Malvern, UK). Colloidal stability was evalu-ated visually according to the presence of sediment on thebottom of the vial.

To determine the molecular weight (Mw) of PDMAAm boundto maghemite, maghemite was first dissolved in concentratedHCl. The resulting dark yellow solution was diluted with waterand PDMAAm separated on a PD 10 desalting column(Amersham Biosciences, Buckinghamshire, UK). Finally, PD-

284 Bioconjugate Chem., Vol. 20, No. 2, 2009 Babic et al.

MAAm was lyophilized and dissolved in acetate buffer (c ∼3mg/mL). The Mw was determined by size-exclusion chroma-tography in an aqueous methanol solution on a Shimadzu HPLCsystem equipped with RI, UV, and multiangle light scatteringDAWN EOS (Wyatt Co., USA) detectors using a mobile phaseconsisting of 20% 0.3 M acetate buffer (pH 6.5; 0.5 g/L NaN3)and 80% methanol (flow rate 0.5 mL/min) with a GPC columnTSKgel G6000PW (300 × 7.8 mm; 15 µm; Tosoh Bioscience,Japan).

The conversion versus DMAAm polymerization time wascalculated from the concentration of monomers during poly-merization monitored after the addition of hydrochinone (50mg/mL) by a GC Perkin-Elmer (Norwalk, CT, USA) with FIDdetector using 30 m length, 0.32 mm inner diameter Rtx-624column with 1.8 µm thick film (Restek, Bellefonte, PA, USA).Calibration was performed using an external standard.

Cell Cultures and Cell Labeling. In vitro cellular uptakeexperiments were performed using rat and human bone marrowmesenchymal stem cells. Rat bone marrow mesenchymal stemcells (rMSCs) were obtained from the tibia and femur of 4-week-old Wistar rats. The ends of the bones were cut, and the marrowwas extruded with Dulbecco’s modified Eagle’s medium(DMEM, PAA Laboratories, Linz, Austria) by using a needleand a syringe (22). Marrow cells were plated in a 75-cm2 tissueculture flask (TPP, Trasadingen, Switzerland) in DMEMmedium containing 10% fetal bovine serum (FBS, PAALaboratories, Linz, Austria), 100 units/mL penicillin (GibcoBRL, Paisley, Scotland), and 0.1 mg/mL streptomycin (GibcoBRL). After 24 h, the nonadherent cells were removed byreplacing the medium.

Human bone marrow mesenchymal stem cells (hMSCs) wereobtained from the bone marrow of healthy donors after obtaininginformed consent. Bone marrow aspirates were diluted inphosphate-buffered saline (PBS) and centrifuged through adensity gradient (Ficoll-Paque Plus, GE Healthcare Life Sci-ences, Vienna, Austria) for 30 min at 1000 g (22, 23). Nucleatedcells from the interface were plated in a 75-cm2 tissue cultureflask in R-modified Eagle’s medium (R-MEM, Gibco BRL)containing 10% FBS, 100 units/mL penicillin, and 0.1 mg/mLstreptomycin. After 48 h, the nonadherent cells were removedby replacing the medium.

Both types of cells were cultured in a humidified 5% CO2

incubator. The medium was replaced every 3 days as the cellsgrew to confluence. The cells were labeled with iron oxidenanoparticles. For cell labeling, either uncoated γ-Fe2O3 nano-particles, the contrast agent Endorem, or PDMAAm-coatedγ-Fe2O3 nanoparticles were used at the same concentration -15µg Fe2O3/mL of media. After 72 h, the contrast agent waswashed out by replacing the culture medium.

Transmission Electron Microscopy (TEM). Nanoparticlelocalization inside the cells was observed by transmissionelectron microscopy. MSCs were incubated with the nanopar-ticles for 72 h, transferred onto polylysine-coated coverslips,fixed in 2.5% glutaraldehyde in 0.1 M Sorensen buffer for 48 hat 4 °C, and stained by 1% osmium tetraoxide in 0.1 M Sorensenbuffer for 2 h. Cells were then dehydrated in ethanol, immersedin propylene oxide, and flat embedded in Epon 812 using gelatincapsules. After polymerization for 72 h at 60 °C, coverslipswere removed by liquid nitrogen. Ultrathin sections of 60 nmwere examined with a Philips Morgagni 268D transmissionelectron microscope (FEI Inc., Hillsboro, OR, USA).

Cell Viability. The viability of cells was analyzed using acolorimetric assay based on the cleavage of the tetrazolium saltWST-1 (4-[3-(4-iodophenyl)-2-(4-nitrophenyl)-2H-5-tetrazolio]-1,3-benzene disulfonate) (Roche Diagnostics, Mannheim, Ger-many) to a highly water-soluble formazan dye by mitochondrialdehydrogenases in viable cells (24). rMSCs or hMSCs (2nd and

5th passages) were plated on 96-well plates (TPP, Trasadingen,Switzerland) at a density of 5 × 103 cells/well. The cells werecultured and labeled with iron oxide nanoparticles as describedabove. Twelve wells of each cell type were labeled withEndorem and 12 wells with PDMAAm-coated maghemitenanoparticles, and 12 wells were not labeled and served ascontrol samples. On the day that nanoparticles were withdrawn(day 3), 10 µL of the WST-1 solution was added to 100 µL ofculture medium per well, and the cells were kept in the incubator(37 °C) for an additional 2 h. The absorbance was measuredusing an ELISA plate reader (Tecan Spectra, Tecan Trading,Switzerland) at a wavelength of 450 nm.

Labeling Efficiency and Staining Intensity. Labeling ef-ficiency was determined by manually counting the number ofPrussian Blue-stained and unstained cells in 96-well plates andmeasuring the staining intensity of Prussian Blue-stained cellscolorimetrically. Twelve optical fields from each plate werescanned using an Axioplan Imaging II microscope at 100×magnification using a 10×/0.75 objective lens, an AxioCamdigital camera, and AxioVision 4 software (microscope setupfrom Zeiss, Oberkochen, Germany). All of the cells in thescanned images were manually labeled as Prussian Blue-stainedor unstained using Jasc Paint Shop Pro 8 (Corel Corporation,Ottawa, Canada). The scanned images with manually labeledcells were processed by the Image analysis toolbox of MATLABsoftware (The MathWorks, Inc., MA, USA). Before the analysis,we validated the colorimetric scale from the image colorscorresponding to the increasing intensities of the Prussian Bluestaining of the cells. Then we processed each Prussian Blue-stained cell in the image and determined the intensity of thecytoplasmic staining as the intensity of the color of thecytoplasm on the scale. As a result, two parameters wereobtained: (i) the presence or absence of a label inside the cellsexpressed as the percentage of labeled cells and (ii) the amountof label inside the cells, which correlates with the intensity ofthe staining.

In Vitro MR Relaxometry and MR Imaging of LabeledCells Suspended in Gelatin. MR relaxometry of the particleswas performed using a Bruker MiniSpec 0.5 T relaxometer anda Bruker BioSpec 4.7 T imager (both Etlingen, Germany). Thelatter apparatus was also used for MR imaging. Two types oflabeled cells, rat and human MSCs, were dispersed in gelatinin the following concentrations: 50, 100, 200, 400, and 800 cells/µL. Control samples containing suspensions of unlabeled cellsand Endorem-labeled cells were measured at the same concen-trations. For T1 relaxometry, the standard saturation recoverysequence was used; T2 relaxometry was performed with a CPMGmultispin-echo sequence. Measured relaxation times wereconverted to relaxation rates and related to cell concentrations.MR imaging was performed with T2-weighted turbo-spin echo(effective echo time TE ) 35 ms, repetition time TR ) 3 s)and T2*-weighted gradient echo (TE ) 5 ms, TR ) 100 ms)sequences that are commonly used in in vivo measurements.

Iron Analysis. The amount of iron in the cells aftermineralization was determined by spectrophotometry. Cell-containing samples were mineralized by the addition of 5 mLof HNO3 and 1 mL of H2O2 in an ETHOS 900 microwavemineralizator (Millestone, Sydney, Australia). Deionized waterwas added to reach a total volume of 100 mL. The iron contentwas determined using a Spectroflame M120S apparatus (Spectro,Littleton, MA, USA) calibrated with a standard Astasol solution(Analytika, Prague, Czech Republic). The measurements wererepeated four times, and the average mean value was determined.

Cell Grafting. The rats (n ) 5) were anesthetized withisoflorane and mounted in a stereotactic frame. Using an aseptictechnique, a burr hole (1 mm) was made in the skull to exposethe dura overlying the cortex. PDMAAm-modified iron oxide-

Maghemite Nanoparticles Bioconjugate Chem., Vol. 20, No. 2, 2009 285

labeled cells or Endorem-labeled cells (5,000 cells suspendedin 5 µL of PBS) were slowly injected intracerebrally into thebrain over a 10-min period using a Hamilton syringe. Unlabeledcells were injected as a control. The opening was closed bybone wax, and the skin was sutured.

MR Imaging. The in vivo MR images were obtained on a4.7 T Bruker imager equipped with a homemade surface coil.The rat was anesthetized by passive inhalation of 1.5-2%isoflorane in air. Breathing was monitored during the measure-

ments. Single sagittal, coronal, and transversal images wereobtained by a fast gradient echo sequence for localizing, thenboth T2-weighted (turbospin echo sequence, TE ) 35 ms, TR) 3 s, number of acquisitions AC ) 8, matrix size 256 × 256,field of view FOV ) 3.5 × 3.5 cm, slice thickness 0.75 mm)and T2*-weighted (gradient echo sequence, TE ) 12 ms, TR )180 ms, number of acquisitions AC ) 48, same geometry) axialimages were acquired.

RESULTS AND DISCUSSION

Preparation of Maghemite. Maghemite (γ-Fe2O3) nanopar-ticles are one of the most commonly used ferric oxide particlesbecause of their simple synthesis and chemical stability. Theiradvantage is that iron participates in human metabolism and isthus well tolerated by the living organism. First, magnetite wasprecipitated; its colloidal stability was achieved by chargesinduced by Fe3+ ions. However, because the above colloid ischemically unstable, undergoing uncontrolled oxidation uponcontact with air, it was controllably chemically oxidized tomaghemite, which shows long-term stability in either alkalineor acidic media. The ultimate advantage of the above procedureconsists in the simplicity of purification by magnetic separation/decantation, which is much more convenient than the commonlyused but more elaborate and time-consuming method ofultracentrifugation.

TEM and SEM images (Figure 1a,b) show agglomeratedmaghemite particles in a dry state. The agglomerates consist ofprimary particles with an average size Dn ) 6.3 nm and apolydispersity index PDI ) 1.33. X-ray diffraction measure-

Figure 1. (a,c) TEM and (b,d) SEM micrographs of (a,b) primary uncoated γ-Fe2O3 particles No. 1 and (c,d) particles coated with PDMAAm bysolution polymerization No. 5.

Figure 2. X-ray diagram of synthesized γ-Fe2O3 nanoparticles. Verticalbars, γ-Fe2O3 standard.


ments of the synthesized particles evidenced the typical spinelstructure of γ-Fe2O3 (Figure 2). With respect to a small size ofthe crystallites, weak superstructural patterns indicated for thestandard material were not detectable. Detailed Mossbauerspectra were published elsewhere (25).

Solution Radical Polymerization of DMAAm in thePresence of Maghemite Nanoparticles. The surface modifica-tion of nanoparticles is a general strategy for enhancing thepermeability of nanoparticle-based therapeutics. To modify theiron oxide surface, the solution radical polymerization ofDMAAm in aqueous ferrofluid was investigated. This couldproduce a dispersion of maghemite nanoparticles stable in amedium suitable for subsequent polymerization of DMAAm,thus facilitating their subsequent incorporation in the polymermicrospheres. At the same time, it provides a model forinvestigating the effect of iron oxide nanoparticles on thepolymerization process. In the system, ACVA was preferredover 2,2′-azobisisobutyronitrile (AIBN) as an initiator becauseits solubility in an aqueous monomer solution is much higherthan that of AIBN. An additional advantage of ACVA consistsin its carboxyl group’s ability to interact with iron oxide. Thepolymerization conditions are summarized in Table 1. The effectof several polymerization parameters, such as the concentrations

of the monomer, the initiator, and the iron oxide in the feed, onparticle size and the amount of PDMAAm bound to the particleswas investigated. For the sake of comparison, the characteristicsof neat (uncoated) maghemite nanoparticles and of nanoparticlescoated with a solution of preprepared PDMAAm are also shownin Table 1.

Conversion Curves. Conversion curves were determined inorder to estimate the effect of maghemite on the polymerizationprocess. The presence of maghemite colloid in the feed affectedthe molecular weight of the resulting PDMAAm compared tosolution radical polymerization without maghemite. The mo-

Table 1. Modification of Maghemite Nanoparticles by the Solution Polymerization of DMAAm

polymerization feed

sampleγ-Fe2O3

(g)DMAAm

(g)ACVA(mg)

colloidalstabilitye Dh

c (nm)bound

PDMAAmd (wt %)

1a 0.5 0 0 - 127 ( 3 02b 0.5 0.5 0 - 161 ( 3 0.213 0.5 0.375 3.8 aggr. n.a.4 0.5 0.5 5 ++ 81.4 ( 0.6 0.75 0.5 1 10 ++ 56.9 ( 0.3 1.26 0.5 1.5 15 + 171 ( 5 1.957 0.5 1 5 ++ 77.8 ( 0.3 1.528 0.5 1 15 + 172 ( 22 1.79 0.5 1 20 aggr. n.a. 2.7410 0.5 1 30 aggr. n.a. 4.9511 0.5 0.375 10 ++ 55 ( 5 1.0312 0.5 0.5 10 ++ 67 ( 3 1.113 0.5 1.5 10 aggr. 328 ( 4 1.0614 0.5 2 10 aggr. 614 ( 12 1.615 1 1 10 ++ 76.6 ( 0.5 2.116 1.5 1 10 ++ 83.5 ( 0.2 0.8817 2 1 10 ++ 75.1 ( 0.1 0.9618 2.5 1 10 ++ 76 ( 0.5 0.74

a Primary uncoated γ-Fe2O3 (control). b PDMAAm solution was added to neat γ-Fe2O3 particles. c Dh, hydrodynamic diameter determined by dynamiclight scattering. d From elemental analysis. e -, poor stability (aggregation within a week); +, high stability (no sediment after 3 months of storage);++, very high stability (no sediment after 6 months of storage).

Figure 3. Conversion curves of DMAAm in solution polymerizationin (∆) the presence and (0) absence of γ-Fe2O3.

Figure 4. ATR FTIR spectra of γ-Fe2O3 nanoparticles before and aftersurface modification with PDMAAm. (a) Before modification, (b)coating by solution polymerization in the presence of γ-Fe2O3 nano-particles No. 3, (c) coating by encapsulation of γ-Fe2O3 nanoparticlesin the polymer No. 2, (b-a) and (c-a) the corresponding differentialspectra of coated and uncoated γ-Fe2O3, and (d) pure PDMAAm.


lecular weight of PDMAAm prepared in the presence ofmaghemite was 717,900, i.e., lower than that prepared in itsabsence (965,200). This can be attributed to the growingpolymer chains probably terminating on the nanoparticle surfaceas documented by the slower consumption of monomers inthe initial and final stages of the polymerization (Figure 3).Because of the termination, PDMAAm can be attached to themaghemite surface. The possible complexation of the ACVAinitiator with maghemite Fe atoms can contribute to additionalDMAAm initiation. At γ-Fe2O3/DMAAm ratios of 0.5/0.375(w/w) and lower, free PDMAAm was present in the mixture

after polymerization. The colloidal stability then decreased withan increasing γ-Fe2O3/DMAAm ratio.

FTIR Spectra. The structure of the surface and the efficiencyof the coating of magnetic nanoparticles were analyzed bysurface-sensitive ATR FTIR spectroscopy. FTIR spectra of ironoxide before modification, after coating by solution radicalpolymerization in the presence of γ-Fe2O3 nanoparticles, aftercoating by encapsulation of maghemite nanoparticles in thepolymer, the corresponding differential spectra of coated anduncoated surfaces, and the spectrum of pure PDMAAm areshown in Figure 4. The spectrum of coating by encapsulationof maghemite γ-Fe2O3 nanoparticles (see corresponding dif-ferential spectrum) is relatively weak and is very close to thespectrum of pure PDMAAm. This signifies that some amountof polymer adheres to the surface of the nanoparticles. Thespectrum of coating by solution radical polymerization in thepresence of maghemite nanoparticles (see corresponding dif-ferential spectrum) is stronger and differs from the spectrum ofpure PDMAAm in some aspects. The band of Amide I (26)observed at 1618 cm-1 is shifted to 1608 cm-1, and the intensityof the band of Amide II observed at 1404 cm-1 increased afterpolymerization. The peaks of CH3 deformation vibration at about1495 and 1354 cm-1 decreased in their relative intensity. Wehypothesize that these changes correspond to the interaction ofthe protonated NH+ groups of PDMAAm with citrates com-plexed on the iron oxide surface. Citrate is adsorbed on the ferricoxide surface via one or two carboxylate moieties, allowing theuse of one of the ungrafted carboxylic acid groups to bind apolycation. This confirms that the PDMAAm shell has ef-fectively formed at the surface of the iron oxide particles.

DLS, TEM, and Elemental Analysis. Particle size and itsdistribution have a strong effect on the physical and biochemicalproperties of colloidal dispersions. Many methods have beendescribed for the measurement of average particle size, such aselectron microscopy, dynamic light scattering, and chromato-graphic methods; the former two techniques were used in thisstudy.

ACVA-initiated solution polymerization of DMAAm in thepresence of maghemite nanoparticles produced a very stablecolloid, with the particles typically in the hydrodynamic size(Dh) range of 50-170 nm. Because DLS provides informationon the hydrodynamic particle size of whole particle clusters,including polymer coating layers and the magnetic core, thesize of the iron oxide core alone was examined by TEM. Figure1c,d shows TEM and SEM images of the obtained DMAAm-coated nanoparticles with an average size Dn ) 7.5 nm and apolydispersity index PDI ) 1.20. A comparison of Figure 1a,band Figure 1c,d reveals that both the shape and the size of theparticles after the coating process effectively did not changefrom those of the uncoated particles. It is worth pointing outthat the particle size determined by TEM was smaller comparedwith the particles size of the same latex obtained from DLSmeasurements (Table 1). These results confirm the presence ofa hydrophilic polymer hairy layer on the particle surface, whichincreases the value of the average hydrodynamic diameter duringthe DLS measurements and which is collapsed onto the particlesurface during the drying of the TEM sample. It should also beremembered that while TEM provides the number-averageparticle size, DLS gives the z-average, which is sensitive tolarge-size particles.

Effect of the Reaction Parameters. The effect of the amountof the monomer in the polymerization feed on colloidal stability,hydrodynamic particle size, and the percentage of PDMAAmbound to the particles (determined from elemental analysis) ata constant initiator concentration (1 wt % in the monomers) isdocumented in Table 1 for experiments Nos. 3-6. Particlesaggregated at low amounts of bound PDMAAm. The percentage

Figure 5. Dependence of the (9) hydrodynamic diameter Dh (a), ([)�-potential, and (]) pH (b) on the ACVA/γ-Fe2O3 ratio.

Figure 6. Dependence of the (a) hydrodynamic particle diameter Dh

and (b) polydispersity, measured by dynamic light scattering, on storagetime; (2) uncoated γ-Fe2O3 No. 1, PDMAAm-coated γ-Fe2O3 ([) No.18, and (9) No. 13.


of bound PDMAAm generally reached up to ca. 2 wt %, whilefree PDMAAm was removed during washing. The moreDMAAm in the polymerization feed, the more PDMAAm wasbound to the particles.

The effect of the amount of initiator in the feed on colloidalstability, hydrodynamic particle diameter, and the percentageof bound PDMAAm is documented in Table 1 for experimentsNos. 5 and 7-10. At <10 mg of ACVA in the polymerizationfeed, the Dh of the resulting particles was smaller than that ofthe primary maghemite particles (Table 1). This can be explainedby the steric repulsion of PDMAAm chains on their surface,which decreases the number of particle associates measured byDLS. At higher ACVA amounts, the stability of the colloidalsystem was lost, and the particles aggregated. Both thehydrodynamic particle size Dh and the percentage of boundPDMAAm determined from elemental analysis increased withincreasing amounts of the initiator in the feed up to 15 mg ofACVA (Nos. 5, 7, and 8 in Table 1). The increasing percentageof bound PDMAAm at higher initiator concentrations can beascribed to a larger number of carboxyl groups available toanchor more PDMAAm chains on the maghemite particle and/or a greater tendency toward particle aggregation. To explainwhy steric stabilization occurred only at concentrations lowerthan 15 mg of ACVA in the feed, ACVA was added to themaghemite colloid, and hydrodynamic diameter Dh, �-potential,and pH were monitored depending on the ACVA/γ-Fe2O3 ratio.

With the ACVA/γ-Fe2O3 ratio increasing from ∼0.02 (w/w),hydrodynamic diameter sharply increased (Figure 5a). This canbe explained by the increase of �-potential from ca. -43 mV,reaching a critical value at -30 mV and accompanied by a steepdecrease of pH till the ACVA/γ-Fe2O3 ratio reached 0.02 (Figure5b). It should be noted that ( 30 mV is the limiting value foreffective electrostatic stabilization. Further ACVA addition thusincreased �-potential up to 13 mV, which was inadequate tostabilize the particles; consequently, they aggregated. Thecoating efficiency of PDMAAm was further deduced from acomparison of maghemite modified by solution radical poly-merization of DMAAm with a sample obtained by the additionof separately prepared PDMAAm to the maghemite colloid. Thedata shown in Table 1 (Nos. 2 and 4) argue in favor of solutionpolymerization.

The effect of the amount of the monomer in the polymeri-zation feed at constant amounts of initiator and maghemite isshown in Table 1 for experiments No. 5 and 11-14. Particleaggregation occurred with amounts of 1.5 g or more DMAAmin the feed, probably because of the undesirable physical cross-linking of PDMAAm chains and the increased viscosity of thereaction mixture. At a constant amount of initiator in the feed,the percentage of PDMAAm bound to the particles did notchange with increasing amounts of DMAAm (with the exceptionof No. 14 in Table 1 because of the high viscosity of themixture), thus documenting the key role of the amount ofinitiator on encapsulation efficiency.

The effect of the amount of γ-Fe2O3 added in the feed underotherwise identical conditions is documented by experimentsNo. 5 and 15-18 in Table 1. While the particle diameter wasalmost constant (within the experimental error), the amount ofbound PDMAAm decreased with increasing amounts of γ-Fe2O3

in the feed, as expected.

Figure 7. (a) Viability of rat (9) and human MSCs (0) labeled with several types of iron oxide nanoparticles. PDMAAm-coated nanoparticles No.7 were used. (b) Labeling efficiency (expressed as the percentage of labeled cells) of rat MSCs from the 1st-5th passage (columns 1-5) labeledwith uncoated γ-Fe2O3 (0), PDMAAm-coated γ-Fe2O3 No. 7 nanoparticles (9), and Endorem (gray square). (c) Distribution of the intensity ofPrussian Blue staining (x-axis) of rMSCs labeled with uncoated γ-Fe2O3 (0), PDMAAm-coated γ-Fe2O3 nanoparticles (9), and Endorem (O). (d)Histograms of the intensity of staining of rat and human MSCs; note that hMSCs (9) are more intensively labeled than rMSCs (0).

Table 2. Cell Labeling Efficiency with Surface-Modified andUnmodified Iron Oxide Nanoparticles

percentage of labeled cells

rMSC hMSC

Endorem 39 ( 6 68 ( 5uncoated γ-Fe2O3 46 ( 5 70 ( 5PDMAAm-coated γ-Fe2O3 59 ( 5 82 ( 5


Colloidal Stability. The long-term colloidal stability ofPDMAAm-coated maghemite nanoparticles is of utmost im-portance for prospective biological applications. The stabilityof some dispersions over time was determined by DLS (Figure6). Almost no increase in the hydrodynamic size and consistentlylow polydispersity over several months were observed in colloidNo. 18, thus documenting its perfect stability in water due tothe presence of the PDMAAm coating. The colloidal stabilityof such samples is denoted as very high in Table 1. The goodcolloidal stability suggests that the stabilization of the maghemite

nanoparticles was primarily dependent on the steric repulsionof the attached hydrophilic PDMAAm chains. In comparison,uncoated nanoparticles No. 1 were unstable. Both the hydro-dynamic diameter and polydispersity of these particles increasedwith time because of aggregation. Similarly, the stability ofnanoparticles No. 13, prepared in the presence of large amountsof monomer, was poor. This can be explained by the highviscosity of the polymerization mixture, as a consequence ofwhich the PDMAAm chains could entangle, resulting in particleaggregation. This is documented by the extremely large sizeand polydispersity determined by DLS in Figure 6a,b.

Viability of Cells. In order to examine the acute toxicity ofPDMAAm-coated maghemite nanoparticles, both rMSCs andhMSCs were incubated for 72 h with sample No. 7, uncoatednanoparticles No. 1, and Endorem (control) at concentrationsof 15 µg γ-Fe2O3/mL, and cell viability was assessed using theWST-1 assay. The viability of both rat and human MSCs labeledwith PDMA nanoparticles did not markedly decrease comparedto that of unlabeled MSCs (control). When mesenchymal stemcells were labeled with Endorem, their viability decreased by32%, and uncoated nanoparticles decreased the viability by 15%(Figure 7a). However, the TEM images showed that the cellslabeled with uncoated nanoparticles were undergoing pro-grammed cell death (see below); therefore, in long-term followup the viability would be affected.

Cell Labeling Efficiency. To investigate the role of thePDMAAm coating and its effect on the internalization of themaghemite nanoparticles by target cells, both hMSCs andrMSCs were incubated with PDMAAm-coated nanoparticles No.7 or Endorem; uncoated maghemite nanoparticles served ascontrols. The uptake of nanoparticles into the cells (expressedas the percentage of labeled cells) was investigated usingPrussian Blue staining. The cellular uptake of particles generallydepends on the cell type and the particle size and surfaceproperties, including surface charge and surface hydrophili-city (27, 28). It also depends on the number of passages of thecells, which is a characteristic physiological cell property. Thelower the passage, the more efficient the cell labeling. The mostintensive staining of any iron oxide nanoparticles was thusobserved for cells from the first and second passages, while theintensity of the blue staining decreased in the fifth passage (only30% of cells were labeled with Endorem). rMSCs cultured withPDMAAm-coated maghemite nanoparticles showed consider-ably higher nanoparticle uptake (59%) than those cultured withEndorem (39%; Table 2). This means that the number of rMSCslabeled by PDMAAm-coated γ-Fe2O3 was more than 50%higher than the number labeled by Endorem. Moreover, thelabeling efficiency (percentage of labeled cells) was stable andwas not dependent on the number of passages (Figure 7b). Evenhigher PDMAAm-nanoparticle uptake was observed in hMSCs(82%; Table 2), whereas Endorem uptake was 68%. The numberof hMSCs labeled by PDMAAm-coated γ-Fe2O3 was thus morethan 20% greater than the number labeled by Endorem.Histograms showing the intensity of Prussian Blue staining,which corresponds to the amount of label inside the cell,revealed that more cells were intensely stained with PDMAAm-coated nanoparticles than with Endorem or maghemite (Figure7c). These results were even more apparent or amplified inhMSCs (Figure 7d). Although a relatively large number of cellswere labeled with uncoated maghemite nanoparticles, theadvantage of the PDMAAm coating consists in the possibilityof its modification, thus introducing desired functional groups.An additional substantial advantage of PDMAAm-coatedγ-Fe2O3 exists in the stability of the formed colloid, in contrastto uncoated γ-Fe2O3 nanoparticles that coagulate after theaddition of the culture media. Coagulation of uncoated maghemitenanoparticles also leads to the attachment of the nanoparticles

Figure 8. TEM micrographs of rMSCs labeled with (a) PDMAAm-coated γ-Fe2O3 nanoparticles, (b) Endorem, and (c) uncoated γ-Fe2O3.Arrows indicate nanoparticles inside the endosomes. A, autophagosome;N, nucleus; n, nucleolus; c1 and c2, cell 1 and cell 2, respectively.Scale bar: a,c, 500 nm; b, 200 nm.


Figure 9. MRI of phantoms containing labeled (a) human and (b) rat mesenchymal stem cells suspended in gelatin measured by a T2*-weightedgradient echo sequence (A-D) and by a T2-weighted turbo-spin echo sequence (E-H). A,E, cells labeled by PDMAA-coated maghemite nanoparticlesNo. 7; B,F, cells labeled by uncoated maghemite nanoparticles No. 1; C,G, cells labeled by Endorem; D,H, unlabeled cells. Each sample (0.5 mL)contained 25,000 cells, yielding on average 1 cell per image voxel. For comparison, an MRI of phantoms containing a high number (400,000) oflabeled human cells is shown in c; the phantoms are in the same order as in a.


to the cell surface. This effect is also responsible for the highlabeling efficiency of uncoated maghemite nanoparticles. It wasdifficult to distinguish using light microscopy whether thenanoparticles were inside the cells or just attached to the cellsurface. Cells with iron oxides attached to the surface are bettertargets for macrophages after transplantation, and iron oxidesincorporated into macrophages can also give a false positivesignal on MRI.

TEM Images. Intracellular uptake was also visualized withTEM as it provides higher resolution than does light microscopy.While these methods cannot provide a quantitative measurementof cellular uptake, they demonstrate the importance of visualiza-tion to determine the distribution of particles on a cellular andsubcellular level. PDMAAm chains are thought to form acomplex with surface charges on the cell surface or else havean affinity with the cell membrane that facilitates endocytosis.A TEM image (Figure 8) shows that the cells internalized thePDMAAm-coated maghemite nanoparticles in large numbersand accumulated them in endosomes. Organelles and cellstructures were not affected by the presence of coated nano-particles inside the cells. However, the majority of cells labeledwith uncoated maghemite were undergoing programmed celldeath, probably by autophagia. The cells contained largeautophagosomes, recognized in micrographs as membrane boundorganelles with other organelles clearly contained within them,together with uncoated nanoparticles (Figure 8b). MSCs stainedwith Endorem showed a heterogeneous distribution of thenanoparticles. Some of the cells contained numerous endosomesfilled with nanoparticles, others only a few or none (Figure 8c).The results obtained from TEM images are in agreement withthose of our colorimetric analysis of the intensity of PrussianBlue staining (Figure 7c) as well as with those obtained fromMR relaxometry (see text below and Table 3).

NMR Relaxometry. The iron oxide concentration of labeledrat and human cells was assessed using MR relaxometry.Relaxation rates measured at 0.5 and 4.7 T are listed in Table3. Relaxation rates r1 and r2 are related to the number of cellsper mL after subtracting the contribution of unlabeled cells. Ata higher field strength (4.7 T), the contribution of all the particlesto total r1 was negligible; the values are therefore not provided.PDMAAm-coated iron oxide-labeled human cells providedsignificantly higher r1 and r2 at both 0.5 and 4.7 T fields thanEndorem- and uncoated-iron-oxide-labeled cells. Large differ-ences were obtained in the relaxation rates of PDMAAm-labeledhuman and rat cells, with r2 several times higher for humancells. PDMAAm-coated iron oxide-labeled rat cells providedhigher relaxation rates than Endorem-labeled ones; however,they were quite comparable to data obtained from uncoated-iron-oxide-labeled cells. Iron analysis proved that the higherrelaxation rates of PDMAAm-coated or uncoated iron-oxide-labeled cells compared with Endorem-labeled ones are causedmainly by higher iron internalization (Table 3). It is interestingto note that human cells prefer PDMAAm-coated nanoparticlesto both Endorem and uncoated ones, whereas in the case of ratcells, uncoated iron particles are internalized slightly better thanPDMAAm-coated ones (both nanoparticles are internalized ata markedly higher rate than Endorem). These results are in goodagreement with the relaxometry results.

MR Imaging of Cells after Nanoparticle Internalizationin Gelatin. Gelatin samples with suspended cells labeled byPDMAAm-coated iron oxide nanoparticles, Endorem, and bareparticles were examined by MRI to evaluate the potential ofPDMAAm-coated iron oxide nanoparticles as a targeted MRcontrast agent. Figure 9 compares MR images of cell phantoms.The T2-and T2*-weighted MR images of phantoms containingcells incubated with PDMAAm-coated and uncoated iron oxidenanoparticles show a significant negative contrast enhancement(signal darkening) over those containing cells labeled withEndorem and unlabeled cells (Figure 9a,b). As the MR imagesare normalized during processing, it is not possible to decide(based on observation with the naked eye) which of the twotypes of nanoparticles provides better contrast. However, theMR image of phantoms containing a high number of cells(Figure 9c) unambiguously proves that cells labeled byPDMAAm-coated iron oxide nanoparticles show significantlyhigher negative contrast enhancement. The results thus confirmthat there is a preferential uptake of PDMAAm-coated iron oxidenanoparicles, i.e., the amount of internalized iron is markedlyhigher (see also Table 3), especially in human cells.

In Vivo MR Imaging. Magnetic labeling and MRI were usedto noninvasively monitor cells injected into rats. PDMAAm-coated iron-oxide- and Endorem-labeled rMSCs (5,000 cells in5 µL of PBS) were injected into the left and right rathemispheres, respectively. Only cells labeled with PDMAAm-coated iron-oxide nanoparicles (left hemisphere) were detected(Figure 10).

CONCLUSIONS

In this article, PDMAAm-coated γ-Fe2O3 nanoparticles wereobtained by the solution radical polymerization of DMAAm inthe presence of maghemite nanoparticles obtained by the

Table 3. Relaxation Rates r1 and r2 (Related to Cell Concentration) of hMSCs and rMSCs Labeled with Endorem (Control), Uncoated(Control) No. 1, and PDMAAm-Coated Maghemite Nanoparticles No. 7, and the Amount of Iron Internalized Inside the Cellsa

hMSCs (s-1/106) cells/mL) rMSC (s-1/106 cells/mL)

r1 (0.5 T) r2 (0.5 T) r2 (4.7 T) Fe (pg/cell) r1 (0.5 T) r2 (0.5 T) r2 (4.7 T) Fe (pg/cell)

Endorem undetectable 0.88 ( 0.34 1.7 ( 0.5 3.2 ( 0.5 0.21 ( 0.19 0.13 ( 0.14 0.72 ( 0.09 undetectableuncoated γ-Fe2O3 0.52 ( 0.09 7.13 ( 0.77 9.8 ( 3.3 17.1 ( 0.3 0.55 ( 0.20 3.67 ( 0.62 4.95 ( 0.17 29.3 ( 4.9PDMAAm-coated γ-Fe2O3 1.19 ( 0.04 27.26 ( 0.24 40.8 ( 1.6 36.9 ( 0.5 0.41 ( 0.13 2.91 ( 0.50 4.18 ( 0.40 23.2 ( 2.9

a Measured at 0.5 and 4.7 T.

Figure 10. T2-weighed MR image of a rat brain with 5,000 rMSCslabeled with PDMAAm-coated nanoparticles No. 7 or Endorem,implanted in the left or right hemisphere, respectively. White and blackarrows indicate the injection sites in the head and the nanoparticles inthe brain, respectively.


chemical coprecipitation of Fe(II) and Fe (III) salts withammonium hydroxide and subsequent oxidation with sodiumhypochlorite. The presence of the PDMAAm coating on themaghemite surface was confirmed by both elemental analysisand ATR FTIR spectroscopy. PDMAAm coating provided gooddispersibility of the magnetic nanoparticles in an aqueousmedium and substantially improved iron oxide colloidal stabilityin the culture medium compared with uncoated maghemite,which immediately precipitated or aggregated within 1 h. A keyparameter determining the amount of PDMAAm bound onmaghemite nanoparticles is the amount of initiator in thepolymerization feed. The amount of the initiator also influencesthe stability of the colloid. A narrow range of reaction conditions(5-15 mg of ACVA in the feed) was found under which particleaggregation was avoided, and at the same time, PDMAAmcoating of the particles was high. DLS data revealed that theparticle sizes were not altered even after several months ofincubation in water. The surface of the particles is the determin-ing factor for the efficiency of cellular uptake. Magnetic cellularlabeling was evaluated by Prussian Blue staining for ironcontent, T2 relaxometry, and MR imaging of labeled cellsuspensions. PDMAAm coating did not adversely affect thepenetration of the iron oxide nanoparticles into cells. After 72 hof culture in media containing 15 µg of iron oxide per 1 mL,the morphology and viability of both rMSCs as well as hMSCswere close to those of unlabeled cells, suggesting the biocom-patibility of the nanoparticles. PDMAAm-coated maghemitenanoparticles turned out to be particularly suitable for labelingrat and human MSCs for in vivo MR cell tracking. ThePDMAAm-coated maghemite nanoparticles were shown to betaken up by the target cells at significantly higher levels thanwere Endorem nanoparticles coated with dextran, thus allowinga minimum amount of imaging agent to be used. In addition,the labeling was not dependent on cell passage number, whichcan be particularly useful for cell expansion. Significant con-trast enhancement was demonstrated by cells labeled withPDMAAm-coated maghemite nanoparticles and implanted ina rat brain over cells labeled with Endorem.

The fact that PDMAAm coating of iron oxide nanoparticlesimproves cellular uptake is, to the best of our knowledge, anew observation yet to be published. In subsequent research,PDMAAm-coated maghemite nanoparticles can be furthermodified by covalent attachment of highly specific biomolecules(antibodies, peptides, etc.) to make labeling more specific. Inpotential applications, cells can be labeled in vitro with thenanoparticles prior to transplantation. After transplantation ofthe labeled cells, in vivo imaging monitors cell migrationthroughout the body. The ability to track the extent of migrationfollowing direct implantation or systemic injection in vivo usingnoninvasive MR imaging techniques is the greatest advantageof the described procedure. Cellular monitoring is particularlyimportant for evaluating cell-based therapies.

ACKNOWLEDGMENT

The financial support of the Center for Cell Therapy andTissue Repair No. 1M0538, the Grant Agency of AS CR (grantsKAN201110651 and KAN200200651), the Grant Agency ofthe Czech Republic (203/09/1242) and the EC-FP6 project DiMI(LSHB-CT-2005-512146) is gratefully acknowledged.

LITERATURE CITED

(1) Lewin, M., Carlesso, N., Tung, C. H., Tang, X. W., Cory, D.,Scadden, D. T., and Weissleder, R. (2000) Tat peptide-derivatizedmagnetic nanoparticles allow in vivo tracking and recovery ofprogenitor cells. Nat. Biotechnol. 18, 410–414.

(2) Modo, M., Hoehn, M., and Bulte, J. W. (2005) Cellular MRimaging. Mol. Imaging 4, 143–64.

(3) Wilhelm, C., Billotey, C., Roger, J., Pons, J. N., Bacri, J. C.,and Gazeau, F. (2003) Intracellular uptake of anionic superpara-magnetic nanoparticles as a function of their surface coating.Biomaterials 24, 1001–1011.

(4) Metz, S., Bonaterra, G., Rudelius, M., Settles, M., Rummeny,E. J., and Daldrup-Link, H. E. (2004) Capacity of humanmonocytes to phagocytose approved iron oxide MR contrastagents in vivo. Eur. Radiol. 14, 1851–1858.

(5) Muller, K., Skepper, J. N., Posfai, M., Trivedi, R., Horwath,S., Corot, C., Lancelot, E., Thompson, P. W., Brown, A. P., andGillard, J. H. (2007) Effect of ultrasmall superparamagnetic ironoxide nanoparticles (Ferumoxtran-10) on human monocyte-macrophages in vitro. Biomaterials 28, 1629–1642.

(6) Hinds, K. A., Hill, J. M., Shapiro, E. M., Laukkanen, M. O.,Silva, A. C., Combs, C. A., Varney, T. R., Balaban, R. S.,Koretsky, A. P., and Dunbar, C. E. (2003) Highly efficientendosomal labeling of progenitor and stem cells with largemagnetic particles allows magnetic resonance imaging of singlecells. Blood 102, 867–872.

(7) Sykova, E., and Jendelova, P. (2007) Migration, fate and invivo imaging of adult stem cells in the CNS. Cell Death Differ.14, 1336–1342.

(8) Veiseh, O., Sun, C., Gunn, J., Kohler, N., Gabikian, P., Lee,D., Bhattarai, N., Ellenbogen, R., Sze, R., Hallahan, A., Olson,J., and Zhang, M. (2005) Optical and MRI multifunctionalnanoprobe for targeting glicomas. Nano Lett. 5, 1003–1008.

(9) Huh, Y. M., Jun, Y. W., Song, H. T., Kim, S., Choi, J. S., Lee,J. H., Yoon, S., Kim, K. S., Shin, J. S., Suh, J. S., and Cheon,J. (2005) In vivo magnetic resonance detection of cancer by usingmultifunctional magnetic nanocrystals. J. Am. Chem. Soc. 127,12387–12391.

(10) Sun, C., Sze, R., and Zhang, M. (2006) Folic acid-PEGconjugated superparamagnetic nanoparticles for targeted cellularuptake and detection by MRI. J. Biomed. Mater. Res. 78A, 550–557.

(11) Tiefenauer, L. X., Kuhne, G., and Andres, R. Y. (1993)Antibody-magnetite nanoparticles: in vitro characterization of apotential tumor-specific contrast agent for magnetic resonanceimaging. Bioconjugate Chem. 4, 347–352.

(12) Kresse, M., Wagner, S., Pfefferer, D., Lawaczek, R., Elste,V., and Semmler, W. (1998) Targeting of ultrasmall superpara-magnetic iron oxide (USPIO) particles to tumor cells in vivo byusing transferring receptor pathways. Magn. Reson. Med. 40,236–242.

(13) Arbab, A. S., Bashaw, L. A., Miller, R. B., Jordan, E. K.,Lewis, B. K., Kalish, H., and Frank, J. A. (2003) Characterizationof biophysical and metabolic properties of cells labeled withsuperparamagnetic iron oxide nanoparticles and transfection agentfor cellular MR imaging. Radiology 229, 838–846.

(14) Gershon, H., Ghirlando, R., Guttman, S. B., and Minsky, A.(1993) Mode of formation and structural features of DNA-cationic liposome complexes used for transfection. Biochemistry32, 7143–7151.

(15) Jesephson, L., Tung, C. H., Moore, A., and Weissleder, R.(1999) Highly-efficient intracellular magnetic labeling with novelsuperparamagnetic-tat peptide conjugates. Bioconjugate Chem.10, 186–171.

(16) Kostura, L., Kraitchman, D. L., Mackay, A. M., Pittenger,M. F., and Bulte, J. W. M. (2004) Feridex labeling of mesen-chymal stem cells inhibits chondrogenesis but not adipogenesisor osteogenesis. NMR Biomed. 17, 513–517.

(17) Strable, E., Bulte, J. W. M., Moskowitz, B., Vivekanandan,K., and Douglas, T. (2001) Synthesis and characterization ofsoluble iron oxide-dendrimer composites. Chem. Mater. 13,2201–2209.

(18) Bulte, J. W., Zhang, S., van Gelderen, P., Herynek, V., Jordan,E. K., Duncan, I. D., and Frank, J. A. (1999) Neurotransplantationof magnetically labeled oligodendrocyte progenitors: Magneticresonance tracking of cell migration and myelination. Proc. Natl.Acad. Sci. U.S.A. 96, 15256–15261.


(19) Horak, D., Babic, M., Jendelova, P., Herynek, V., Trchova,M., Pientka, Z., Pollert, E., Hajek, M., and Sykova, E. (2007)D-mannose-modified iron oxide nanoparticles for stem celllabeling. Bioconjugate Chem. 18, 635–644.

(20) Babic, M., Horak, D., Trchova, M., Jendelova, P., Glogarova,K., Lesny, P., Herynek, V., Hajek, M., and Sykova, E. (2008)Poly(L-lysine)-modified iron oxide nanoparticles for stem celllabeling. Bioconjugate Chem. 19, 740–750.

(21) Orakdogen, N., and Okay, O. (2006) Reentrant conformationtransition in poly(N,N-dimethylacrylamide) hydrogels in water-organic solvent mixtures. Polymer 47, 561–568.

(22) Azizi, S. A., Stokes, D., Augelli, B. J., DiGirolamo, C., andProckop, D. J. (1998) Engraftment and migration of human bonemarrow stromal cells implanted in the brains of albino rats:Similarities to astrocyte grafts. Proc. Natl. Acad. Sci. U.S.A. 95,3908–3913.

(23) Pittenger, M. F., Mackay, A. M., Beck, S. C., Jaiswal, R. K.,Douglas, R., Mosca, J. D., Moorman, M. A., Simonetti, D. W.,Craig, S., and Marshak, D. R. (1999) Multilineage potential ofadult human mesenchymal stem cells. Science 284, 143–147.

(24) Ishiyama, M., Shiga, M., Sasamoto, K., Mizoguchi, M., andHe, P. (1993) A new sulfonated tetrazolium salt that produces a

highly water-soluble formazan dye. Chem. Pharm. Bull. 41,1118–1122.

(25) Zaveta, K., Kancık, A., Marysko, M., Pollert, E., and Horak,D. (2006) Superparamagnetic properties of γ-Fe2O3 particles:Mossbauer spectroscopy and d.c. magnetic measurements. Czech.J. Phys. 56 (Suppl. E), 83–91.

(26) Socrates, G. (2001) In Infrared and Raman CharacteristicGroup Frequencies, pp 143-145, Wiley, New York.

(27) Sun, R., Dittrich, J., Le-Huu, M., Mueller, M. M., Bedke, J.,Kartenbeck, J., Lehman, J. K., Krueger, R., Bock, M., Huss, R.,Seliger, C., Grone, H. J., Misselwitz, B., Semmler, W., andKiessling, F. (2005) Physical and biological characterization ofsuperparamagnetic iron oxide- and ultrasmall superparamagneticiron oxide-labeled cells: A comparison. InVest. Radiol. 40, 504–513.

(28) Bentzen, E. L., Tomlinson, I. D., Mason, J., Gresch, P.,Warnement, M. R., Wright, D., Sanders-Bush, E., Blakely, R.,and Rosenthal, S. (2005) Surface modification to reduce non-specific binding of quantum dots in live cell assays. BioconjugateChem. 16, 1488–1494.

BC800373X


Poly( N , N -dimethylacrylamide)-Coated Maghemite Nanoparticles for Stem Cell Labeling

Documents

Transcript of Poly( N , N -dimethylacrylamide)-Coated Maghemite Nanoparticles for Stem Cell Labeling